1. Field of the Invention
The present invention relates to a semiconductor device and a method for manufacturing the same, and more particularly to a semiconductor device having a solder ball formed on an electrode pad and a method for manufacturing the same semiconductor device.
2. Description of the Related Art
In recent years, the miniaturization in size and reduction in cost of electronic appliances such as mobile telephones and notebook-sized personal computers have advanced, and a semiconductor device such as a large-scale integrated circuit and the like (hereinafter referred to as an LSI) to be built in these electronic appliances is required to be more miniaturized in size and more reduced in cost.
In order to meet such a demand as described above, a flip chip ball grid array (FCBGA) method has appeared which fixes solder balls on a plurality of electrode pads on an LSI, connects the solder balls directly to the corresponding electrodes of a wiring board and thereby joins the LSI electrically and mechanically to the wiring board. In an LSI using this method, the surface layer of an electrode pad is formed out of copper (Cu) in order to improve the affinity of a solder ball for the electrode pad and make the connectivity better.
A conventional semiconductor device using an FCBGA method is disclosed in Japanese Patent Laid-Open Publication No. Hei 10-261,642. FIG. 19 is a magnified sectional view showing a solder ball forming portion of a semiconductor device described in this reference.
This semiconductor device comprises a bonding pad 41 formed on a semiconductor substrate 31, a passivation film 33 formed around the bonding pad 41, a barrier film 38 formed on the passivation film 33 and the bonding pad 41, an electrode pad 39 formed on the barrier film 38 in the same flat shape as the barrier film 38, and a solder ball 34 fixed on the electrode pad 39. The barrier film 38 is formed out of Ti, TiW, Cr or TiN, and the electrode pad 39 is formed out of Cu for the above-mentioned reason.
In such a conventional semiconductor device manufacturing method as described above, since a solder ball 34 stops at the outer edge portion of a barrier film 38 as covering the whole of an electrode pad 39 when it is reflowed and is fixed on the electrode pad 39 nearly in the shape of a sphere due to its surface tension, the solder ball 34 is brought into contact with the outer edge portions of both the electrode pad 39 and the barrier film 38. Namely, since solder is good in wettability with Cu and is poor in wettability with Ti (titanium), solder goes around onto the side surface of the electrode pad 39 made of Cu and stops going around at the boundary between the electrode pad 39 made of Cu and the barrier film 38 made of Ti.
Due to heating for the solder reflow, as shown in FIG. 20, tin (Sn) atoms 43 thermally diffuse and move from the interface between the solder ball 34 and the electrode pad 39 into the electrode pad 39. The inventors have found that at this time, since the quantity of movement of Sn atoms 43 in the interface between the electrode pad 39 and the barrier film 38 is larger than the quantity of movement of them in the interface between the solder ball 34 and the electrode pad 39, the Sn atoms 43 have moved to a considerably deep interior from the side surface of the electrode pad 39.
And when patterning the electrode pad 39 by means of etching and the like, its outer edge portion is sometimes not vertical, but is slanted. As shown by a dashed line in FIG. 20, which is a magnified view of a joint portion of FIG. 19, when the outer edge portion of the electrode pad 39 is formed in a tapered shape, Sn atoms 43 of an Sn component thermally diffuse from the tapered portion of the electrode pad 39 and the Sn atoms 43 easily reach the boundary between the electrode pad 39 and the barrier film 38.
For such a reason as described above, the inventors have found also that Sn atoms 43 existing in the interface between the electrode pad 39 and the barrier film 38 degrade the adhesion of the electrode pad 39 to the barrier film 38.
Generally, a solder ball is fixed on a wiring board (not illustrated) by being reflowed at a temperature of 355 to 365 C. At this time, a strain is generated by the difference in thermal expansion coefficient between a semiconductor chip including a semiconductor substrate and a wiring board. At this time, when there is said degradation in adhesion, as shown in FIG. 20, exfoliation 42 occurs at the interface between the electrode pad 39 and the barrier film 38. The exfoliation 42 triggers exfoliation of the solder ball 34 to cause a continuity failure, or triggers the increase of electric resistance to cause degradation in the manufacturing yield rate.
In consideration of the above, an object of the present invention is to provide a semiconductor device and a method for manufacturing the same which can prevent an electrode pad from exfoliating from a barrier film, improve the manufacturing yield rate, and improve the reliability of connection of a semiconductor device with a wiring board even if a strain caused by the difference in thermal expansion coefficient between the semiconductor device and the wiring board acts on a solder ball.
In order to attain the above object, according to a first aspect of the present invention, there is provided a semiconductor device comprising a barrier film, an electrode pad directly contacting the barrier film, a solder ball directly contacting the electrode pad, and a side wall film separating the solder ball from a boundary between the barrier film and the electrode pad.
According to a second aspect of the present invention, there is provided a semiconductor device comprising a wiring layer, an insulating layer on the wiring layer, an opening formed in the insulating layer to expose an upper surface of the wiring layer, a barrier film formed on an inner surface of the opening and on the insulating; an electrode pad formed on the barrier film, a side wall film formed on a side surface of the barrier film and a side surface of the electrode pad covering a boundary between the barrier film and the electrode pad, and a solder ball formed on the electrode pad.
According to a third aspect of the present invention, there is provided a method of manufacturing a semiconductor device comprising the steps of forming an insulating layer on a wiring layer, forming an opening in the insulating layer, the opening exposing an upper surface of the wiring layer, forming a first conductive film on an inner surface of the opening and on the insulating layer, forming a second conductive film on the first conductive film, patterning the first and second conductive films to form a barrier film on the wiring layer and an electrode pad on the barrier film, forming a side wall film covering a boundary between the barrier film and the electrode pad on patterned edges of the barrier film and the electrode pad, and forming a solder ball on the electrode pad.
According to the present invention, based on the above constitution, since the solder ball does not contact the boundary between the electrode pad and the barrier film, it is possible to prevent a component of the solder ball from diffusing into an interface of the electrode pad and the barrier film. Therefore, it is possible to prevent the electrode pad from exfoliating from the barrier film.